Berry, William, John Freeman and Brian Job. “Using Markov Models for Analyzing Perceptual Frameworks.” Political Methodology, 9, no. 3 (1983): 249-84.

Much effort has been devoted to modeling the behavioral interaction between nation-states. However, comparatively little of this work has explicitly recognized the important link between decision-makers’ perceptions and nation-state behavior. Recognizing the difficulty in obtaining reliable information about decision-maker perceptions, some students of international conflict have advocated the use of Markov chains to infer perceptions from observed sequences of behavior.

This article develops a general Markov model for studying the perception-behavior linkage. This general model subsumes such prototype frameworks as the “closed” and “open” models that Zinnes (1968) and others have employed, while allowing for the investigation of a variety of alternative perceptual frameworks. Applying new methodology, we attempt to discern which of eight different perceptual models best accounts for the Sino-Indian conflict of 1959-1964. In the course of this application, errors in an extant Markov analysis of the same conflict are uncovered, and several methodological issues regarding the usefulness of Markov analysis are illuminated.